OGG-01028问题处理记录
参考：
http://blog.itpub.net/26277071/viewspace-713360/
http://oracle11ggotchas.com/articles/TheImportanceofBoundedRecovery.htm




1、环境：
OGG版本：11.1.1
OS:suse 10 sp4 x64
DB:oracle 10.2.0.5 RAC x64


2、原因：由于读取不到归档日志，导致抽取失败，重新挂载归档日志后，抽取到下一个日志时报OGG-01028错误，错误内容如下：


2017-12-14 16:33:26  ERROR   OGG-01028  Record position (SeqNo: 11845, RBA: 2064, SCN: 3628.259323944 (155824006
73832)) is beyond end of recovery (SeqNo: 11844, RBA: 40830128, SCN: 3628.255608022 (15582396957910), Timestamp:
 2017-12-09 16:18:58.000000).


3、查资料分析：
Ans:- It the hitting with the bug 12330566 and the fix will be available only with 11.2 version only. Till that time you need to wait and the workaround is to use BR BROFF refer doc:- OGG Extract abends with OGG-01028 Record position is beyond end of recovery (Doc ID 1335470.1)


4、问题解决执行过程：
查看BR文件
oracle@ptdb02:/ogg/11.1/BR/EXT_S1> pwd
/ogg/11.1/BR/EXT_S1
oracle@ptdb02:/ogg/11.1/BR/EXT_S1> ls -l
total 40
-rw-rw---- 1 oracle oinstall 65536 2017-12-09 16:42 CP.EXT_S1.000004442
-rw-rw---- 1 oracle oinstall 65536 2017-04-12 18:51 PO.EXT_S1.0140659.0000000
drwxrwxr-x 2 oracle oinstall  4096 2014-12-12 18:35 stale


编辑抽取进程参数
edit param ext_s1
参数文件添加
BR BROFF


重启抽取进程
start ext_s1


报如下错误
2017-12-14 20:13:28  ERROR   OGG-00446  Could not find archived log for sequence 7234 thread 1 under default des
tinations SQL <select to_char(first_time, 'YYYY-MM-DD') from gv$log_history where sequence# = :ora_seqno and thr
ead# = :ora_thead>, error retrieving redo file name for sequence 7234, archived = 1, use_alternate = 0Not able t
o establish initial position for sequence 7234, rba 54123536.
2017-12-14 20:13:28  ERROR   OGG-01668  PROCESS ABENDING.


查看归档日志：1节点保存归档日志最老为7663，时间为2017-5-1日的；


指定时间点，重新抽取；
1.       First of all, the Extract process was stopped with the force option.
GGSCI 4> send extract ext_s1, forcestop
2.       The start position of the Extract process was altered to the beginning of the long running transaction.
GGSCI 5> alter extract ext_s1, begin 2017-12-09 16:18:58
3.       The extract process was started.
GGSCI 4> start extract ext_s1
4.       Sure enough, the Extract process was reinitialized and continued to process the backlog.
GGSCI 5> info ext_s1, detail


查看状态正常。